All tasks / Models
Step-Audio R1.1
StepFun's open-weights 33B audio reasoning model that can run as a realtime speech-to-speech agent.
- StepFun
- Open weights
- Released 14 Jan 2026
Where it ranks
- –Voice agentsModels · Local · no result on this task’s benchmarks yet
Facts
- Vendor
- StepFun
- Family
- Step-Audio R1
- Released
- 14 Jan 2026
- Weights
- Open: download and run it yourself
- License
- Apache-2.0
- Parameters
- 33.5B
- GPU memory to run locally
- ~74 GB (Derived: about 33.50B parameters is roughly 67.0 GB of bf16 weights plus runtime overhead; needs a datacenter GPU or multi-GPU setup at bf16; not an official requirement.)
- Access
- Free (open weights)
- Last verified
- 19 Oct 2026
Benchmark results
| Benchmark | Result | Entry on the leaderboard | As of | Source |
|---|---|---|---|---|
| Artificial Analysis Big Bench Audio (speech reasoning) | 97.6% | Step-Audio R1.1 (Realtime)Big Bench Audio speech-reasoning accuracy; embedded bbaScore 0.976 converted to percent; no last-updated date shown. | 19 Oct 2026 | artificialanalysis.ai |
Where you can use it
No tracked tool lists this model yet; as open weights it runs in local runtimes.